The Technology Industry Is Stumbling Down The Path To Becoming A Proper Supply Chain

The technology industry has always been immature compared with the manufacturing, automotive, telecommunications, and oil and gas industries. In those industries, tier-one suppliers are deeply embedded in the operations of the end producer, tier-two suppliers are embedded in the operations of the tier-one suppliers, and so on. Boeing planes would fall out of the sky without deep integration between its engineering and manufacturing teams and those of its jet engine partners General Electric or Rolls-Royce. And those jet engines would fail if the coatings used in turbine blades became brittle and cracked. (This one I know from my father’s pioneering work in sputtering.) The entire supply chain must work together to keep airplanes flying. They are embedded in each other’s success (and in your life).

The tech sector has always been different. The chaos of innovation prevented the stability of mature technology partners from forming — that is, until the cloud kicked into gear. At that point, investment in compute, storage, and software created consolidated global giants. The rest of the sector has remained chaotic, leaving enterprises to stitch together their own collection of suppliers. But the signals of a maturing technology supply chain are getting louder:

Hyperscalers began selling bigger relationships into corporate sourcing, not just IT. Their pivot to consumption-based pricing created a tighter link between spend and value (consumption being a decent proxy for results in an effective deployment). The commitments between customers and suppliers grew stronger, with more co-innovation, more executive relationships, and more money spent. In one deal, Microsoft and Humana CEOs announced a partnership to use Microsoft technology inside Humana’s aging population service portfolio.
Software companies are pushed by AI to price performance and not seats. For decades, the software industry lazily pushed seat-based pricing, locking customers into software deployments that may or may not deliver something valuable. This is little better than the old days of selling shelfware. AI agents are changing the software power dynamic, pushing both suppliers and customers toward results-based pricing. This is a painful reshuffling of investment priorities and the software ecosystem, but it’s long overdue and another signal of a maturing tech supply chain.
Palantir started the craze of forward-deployed engineers to help clients make sense of its AI machinery. Now every model builder, hyperscaler, and SaaS player is following suit. Are they giving away labor? Or helping you deploy their systems? Yes to both. Amazon Web Services’ announcement of a thousand forward-deployed engineers is just the latest announcement, more notable for its scale than its innovation, to push agents into customer workflows. You’re already locked into Amazon Web Services; you may as well get the most of it you can. Microsoft scurried to make its own announcement in the Microsoft Frontier Company for the same purpose. OpenAI sought PE money to build a captive OpenAI Deployment Company.
Enterprises are outsourcing more capabilities that are mission-critical but not differentiating. This is leading to a vast expansion of managed services for functions such as supply chain management, treasury functions, claims processing, and more. AI-powered automation creates incentives for enterprises to offload internal operations to service provider partners. EY, for example, builds on P&G’s supply chain operations IP to help other manufacturing clients upgrade their operations. The IP is bundled into the service contract and P&G gets a royalty.
Service providers are committing more resources to their largest clients. This has been going on for decades but is now accelerating. Accenture reports the number of $100 million deals it does every quarter to signal that those customers matter more. The scale and footprint of these relationships create tighter linkages, more interdependencies, and higher incentives for both parties to work together. Other service providers — Infosys, TCS, Wipro, Capgemini — are following the same playbook.
IT procurement is beginning to adopt the patterns of corporate sourcing, moving beyond seat-based software pricing and time-based service pricing to ask for pricing models more tightly linked to performance or results. Fixed price for technology services or consulting is a reasonable next step beyond time-and-materials pricing, but the future service provider relationship will be more complex performance-based pricing.

How should CIOs deal with this shift? It’s still early going in the industry’s great awakening. Despite these promising signals, tech and service providers have yet to fully embrace their embedded business models, and they and you are still working out what the new relationship should be. CIOs must start today, though, and:

Start thinking of technology suppliers as embedded. It probably means asking your corporate supply chain teams to help you negotiate the terms. They’ll have a perspective on the big picture of risks, financing, and commitment. For example, suppliers like Palantir and Amazon may wish to give you forward-deployed engineers for a few months. You may need them for a few years to make sure the technology works in your business. Push your suppliers to be more embedded in your operation, not just your design and build challenge.
Don’t throw out your consulting and systems integrator partner. Even if you push a tech supplier to leave their forward-deployed engineers in place, that’s still only a piece of the solution. To do a solid Salesforce consolidation, for example, you may still need a service provider to help with the systems integration, data readiness, and change management and perhaps the ongoing technical operations. Force your tech supplier and your service partners to work together by contractual design, not alliance goodwill.
Plan for the inevitable lock-in. You can’t dislodge a technology supplier that’s embedded in your business operation without massive pain. But just as in the aircraft industry, you can’t replace a GE engine with a Rolls-Royce engine without rewiring the entire plane. The same is true with your embedded tech partners. Be careful not to scale your commitments until you have years, maybe a decade, of trust and experience with them.